| <<<Back 1 day (to 2014/03/27) | 2014/03/28 |
Srinivasa_ | Hi all | 09:16.19 |
| what is the algorithm to detect tables in a pdf for extraction? | 09:16.34 |
kens | Do you mean 'How does MuPDF identify tables when performing text extraction' ? | 09:17.17 |
Robin_Watts | If so the answer is: It's an undocumented algorithm using HairyHeuristics(TM). | 09:27.07 |
| Morning Jogux | 09:36.38 |
jogux_mac | morning all | 09:41.34 |
kens | Morning | 09:41.40 |
kens2 | Morning Michael | 14:08.18 |
Robin_Watts | paulgardiner, jogux_mac: I've just updated robin/master with (as well as all the stuff from yesterday), a new version of the solution. | 14:09.25 |
| If you trigger builds from the solution, they now build into a top level 'genroot' directory. | 14:09.47 |
| and all the system vars etc take care of themselves. | 14:10.25 |
| All the commits on robin/master should be good to go now, except for the final few with the new fonts. I need to rejig those so that the fonts replace the old ones. | 14:11.02 |
| I'll do that after lunch. | 14:12.02 |
spanners | gen..root | 14:14.29 |
spanners | crumbles into dust | 14:14.33 |
paulgardiner | I've been slowly working through them. I wondered, should we have updated winbuild.py rather than introducing android.py? (on the other hand it works). | 14:14.56 |
henrys | paulgardiner, Robin_Watts et al, miles brought up the idea of a free app with paid upgrade for editing. Is that a huge engineering ordeal? | 14:30.17 |
| I assume everyone heard the MS Office announcement | 14:30.58 |
paulgardiner | Probably not that horrendous | 14:31.17 |
| Not sure what's involved in making the upgrade process streamlined | 14:32.42 |
henrys | paulgardiner: I've heard doing buy in's on the apple store was very complicated but it is supposed to be better now. | 14:32.56 |
paulgardiner | So we are pretty sure that customers that view only are a small proportion? I guess there are other free options out there for that. | 14:34.53 |
| He he, better: let them edit but don't let them save :-) | 14:35.24 |
henrys | paulgardiner: it's in response to MS office | 14:35.27 |
paulgardiner | Oh right. I hadn't heard, but now I can guess | 14:35.51 |
henrys | paulgardiner: I'll bring a newspaper to the next staff meeting ;-) | 14:36.12 |
kens2 | I read the announcement last night | 14:36.23 |
henrys | I wonder if they made a special deal with apple - they certainly aren't' going to give 30 % of the subscription to apple - the usual apple store extortion. | 14:37.34 |
paulgardiner | Anyone got a url? | 14:37.43 |
henrys | http://arstechnica.com/information-technology/2013/02/microsoft-office-2013-pro-released-to-the-masses-office-365-updated/ | 14:38.52 |
| sorry wrong linke | 14:39.38 |
kens2 | http://www.theregister.co.uk/2014/03/28/microsoft_already_pointing_out_bugs_in_office_for_ipad/ | 14:39.39 |
| http://www.theregister.co.uk/2014/03/27/microsoft_office_apps_ipad/ | 14:40.22 |
henrys | http://arstechnica.com/information-technology/2014/03/microsoft-steps-up-its-mobile-management-game-with-enterprise-mobility-suite/ | 14:41.27 |
| paulgardiner: I kind of like ars coverage - and they have good feedback | 14:42.17 |
| I don't know miles continues to be positive but it seems like app store sales would dry up for clones. | 14:44.48 |
kens2 | would assume so | 14:45.55 |
Robin_Watts | paulgardiner: I thought that winbuild.py was for builds *targetted* at windows ? | 14:46.51 |
| not builds *hosted* on windows. | 14:47.13 |
| tgvbuild being for builds targetted on tgv, etc. | 14:47.38 |
paulgardiner | Ah, that may well be right. Of course tgvbuild is the one that targets many OS's | 14:49.58 |
| Yes good. Ignore that. :-) | 14:51.31 |
| robin_watts Does ndk-build add the _NDK_ define to cflags? | 14:52.19 |
Robin_Watts | possibly. I'd have to check. let me run gitk. | 14:52.51 |
| 'NDK_BUILD_DEFINES':' -D_NDK_ ', | 14:54.03 |
paulgardiner | Where's that then? | 14:54.37 |
Robin_Watts | COMMON_FLAGS += os.environ['NDK_BUILD_DEFINES'] | 14:54.52 |
| In the very first commit on my repo. | 14:55.01 |
paulgardiner | Ah, of course. I looked in the tool configs, but it can't go in there because that is more general than just android | 14:57.31 |
| I guess my initial adding pdf export to the test shell is subsumed, or at least no longer needed? | 15:00.59 |
Robin_Watts | paulgardiner: A modified form of that commit is in my set, yes. | 15:01.24 |
| I rolled all the fixes/tweaks to it back into it. | 15:01.42 |
paulgardiner | good. | 15:01.57 |
| 59a9921 removes debug from linkflags. Was that just there wrongly, or was it needed only for the separate manifests? | 15:04.06 |
Robin_Watts | paulgardiner: I think it was a mistake in the files we inherited. | 15:05.20 |
paulgardiner | There's commented out code added in that one, possibly intentionally of course | 15:05.48 |
Robin_Watts | In the tools.py file? | 15:06.39 |
| That's a slip on my part. I will fix it. | 15:06.45 |
paulgardiner | not added, removed by commenting I mean | 15:06.47 |
Robin_Watts | yes, I'd failed to tidy up once I got it working. | 15:07.23 |
jogux_mac | henrys : Apple are claiming MS are giving them the 30% cut. | 15:07.30 |
Robin_Watts | so MS made it 30% more expensive? :) | 15:08.14 |
jogux_mac | henrys : as to editting as an in-app-purchaseâ¦. itâs certainly doable. thereâs an effort per-appstore (ios, google play, amazon, â¦), plus some changes to the core app to enable/disable editting. | 15:09.29 |
| robin_watts : donât think so. I think thereâs another apple rule that IAP must not be more expensive than the price for the same item elsewhere. | 15:09.53 |
paulgardiner | Releasing a lite version would be a slightly easier alternative | 15:10.23 |
Robin_Watts | jogux_mac: Where else is the ipad version available? | 15:10.36 |
jogux_mac | robin_watts : you donât buy the ipad version; you buy a microsoft 365 subscription (which covers you on all OSes) | 15:11.01 |
Robin_Watts | paulgardiner: Updated commits on line. | 15:11.20 |
jogux_mac | already has an office365 sub, so gets automatic access to the ios version | 15:11.24 |
Robin_Watts | jogux_mac: Ah! | 15:11.26 |
| And what does a 365 sub go for these days ? | 15:11.44 |
jogux_mac | $100 per year | 15:11.50 |
| btw: it was less publised, but office for iphone is now free (for âhomeâ use) I believe. | 15:12.27 |
Robin_Watts | I contend that the kind of person thinking of spending $10 on Smart Office is a different beast to those that would spend $100 a year on MS office. | 15:13.05 |
henrys | paulgardiner: it's probably good planning to have a "no edit" build anyway ⦠| 15:13.18 |
paulgardiner | robin_watts my thoughts too, but free for home use has to be a killer | 15:13.35 |
Robin_Watts | jogux_mac: Don't we already have a Smart Office 'Lite' ? | 15:13.38 |
jogux_mac | henrys : a âno editâ build would likely be a completely different best from a âeditting enabled at runtimeâ build (perhaps) | 15:13.50 |
| Apple get /very/ picky about âLiteâ apps. | 15:14.01 |
Robin_Watts | oh, the Lite app is android only maybe. | 15:14.17 |
jogux_mac | robin_watts : that rings bells. probably for the reason I just mentioned ;-) | 15:14.45 |
| iPhone/iPads come with built in office & pdf /viewing/ | 15:15.01 |
paulgardiner | So users are pretty unlikely to even look for an alternative unless they wish to edit | 15:16.39 |
| ... I'd have thought. | 15:16.55 |
ray_laptop | jogux_mac: what is causing all of those strange characters in your IRC messages? They make the irclogs really hard to read | 15:17.27 |
| jogux_mac: is that Gaelic or something ;-) | 15:17.46 |
kens | suspecta they are proper quotes | 15:18.10 |
ray_laptop | jogux_mac: seems to show up when you are typing "don't" | 15:18.23 |
Robin_Watts | ray_laptop: They look fine in Chatzilla, it's just the logs. | 15:18.24 |
kens | Probably Unicode or something | 15:18.25 |
Robin_Watts | I suspect it's utf-8, yeah. | 15:18.30 |
jogux_mac | ray_laptop : I blame whoever suggested colloquy ;) | 15:18.36 |
| but, yeah, I think itâs a bit in the logs. | 15:18.44 |
tor8 | utf-8 encoded ellipsis characters | 15:18.45 |
ray_laptop | jogux_mac: chatzilla is quite nicce | 15:18.53 |
Robin_Watts | chatzilla certainly doesn't vomit utf-8 into the logs :) | 15:19.21 |
tor8 | jogux_mac: that ellipsis thing is a standard feature of macosx text-input widgets | 15:19.23 |
jogux_mac | ray_laptop : Robin originally suggested that to me, but I instantly found half a dozen ways to crash firefox (and hence chatzilla) :-) | 15:19.24 |
kens | wishes to make it clear he didn't *recommend* colloquy | 15:19.30 |
henrys | paulgardiner: we certainly won't get any customers that just want to view and now we've just lost anyone with a subscription, anyone who cares about writing and insists on "compatibility" will go subscription. I don't see a lot left over after that but I hope I'm wrong. | 15:19.38 |
ray_laptop | donâÃâ¬Ãâ¢t is much harder to read. | 15:19.45 |
tor8 | actually, it's not utf-8 ... it's mangled utf-8 | 15:20.08 |
jogux_mac | kens : I fail to recall who did :-) | 15:20.08 |
tor8 | utf-8 parsed as latin-1 re-emitted as utf-8 | 15:20.17 |
jogux_mac | there seems to be an option to limit it to ascii. let me try that. | 15:20.17 |
kens | I mentioned that henrys uses it :-) | 15:20.20 |
Robin_Watts | Jogux, paulgardiner: So, I'm rejigging the fonts. Putting the modified versions into epage seems sensible enough. | 15:20.31 |
tor8 | jogux_mac: or turn off the "autocomplete abbreviations" thing | 15:20.41 |
Robin_Watts | But what do I do with the stuff in resources? | 15:20.41 |
paulgardiner | henrys: yes my feeling too | 15:20.44 |
jogux_mac | kens : :-) | 15:20.45 |
ray_laptop | well, maybe henrys can divulge his settings to jogux_mac | 15:20.50 |
tor8 | it turns TM into the trademark, etc | 15:20.50 |
jogux_mac | robin_watts : I donât really understand what youâve done | 15:20.58 |
| donât blahblah does still break the logs? | 15:21.20 |
kens | Still weird symbols | 15:21.25 |
jogux_mac | letâs try a restart | 15:21.36 |
henrys | ray_laptop: I didn't change anything other than to mute sounds. | 15:22.01 |
ray_laptop | henrys: probably a locale setting or something | 15:22.18 |
Robin_Watts | jogux_mac: I take the metrics from the official MS fonts. I take the outlines from the EFF ones. I smash them together, and I get modified fonts out with the EFF outlines and the MS metrics. | 15:22.18 |
jogux_mac | donât blahblahblah | 15:22.27 |
| henrys : it will be interesting to see what the compatibility is like between office ipad and desktop office. | 15:23.27 |
Robin_Watts | Jogux, paulgardiner: So, for the resources do I: | 15:24.01 |
| 1) commit my modified versions in a "modified"directory alongside the originals. | 15:24.24 |
jogux_mac | henrys : I believe a lot of the recent money has come from customers (OEMs/integrators/etc) who want in-app solutions; they donât want to pass a document out to an external app to edit. | 15:24.44 |
Robin_Watts | 2) move the originals into an "originals" directory, and put the modified ones where the originals were. | 15:24.46 |
| 3) replace the originals with my bastardised versions. | 15:25.05 |
jogux_mac | you didnât change resources at all yet right? | 15:25.16 |
spanners | Robin_Watts: How are you modifying them? | 15:25.20 |
Robin_Watts | spanners: I take the metrics from the official MS fonts. I take the outlines from the EFF ones. I smash them together, and I get modified fonts out with the EFF outlines and the MS metrics. | 15:25.38 |
| jogux_mac: So far, resources is entirely unchanged. | 15:25.50 |
spanners | It's the 'smash' I'm intrigued by. | 15:25.56 |
Robin_Watts | spanners: craptastic tool I wrote. | 15:26.07 |
spanners | Right. Because when I did font modifications I used FontTools/TTX which I found to be the most robust and cleanest way of doing it. | 15:26.37 |
| That's the tool which converts TTF/T1 fonts into XML. | 15:26.52 |
| (and back) | 15:26.56 |
Robin_Watts | spanners: Interesting. I wish I'd known about that :) | 15:27.07 |
| I work on the TTFs directly. | 15:27.15 |
spanners | That's why I'm lurking. | 15:27.18 |
| Looks like there's a more up-to-date fork: https://github.com/behdad/fonttools/ | 15:28.45 |
henrys | jogux_mac: agreed, we were hoping for some app store sales though. We'll see. | 15:31.35 |
jogux_mac | nods at henrys; I think at Robin says the people who pay 7 quid for smartoffice probably doesnât have a huge intersection with the people that pay 70+quid per year for MS stuff. but we shall see. I need to have a proper play with the ipad stuff (couldnât remember my office 365 password last night :-) ) | 15:57.47 |
Robin_Watts | spanners: Thanks. I'll have a look at that if I ever have to retread this ground. | 15:58.42 |
jogux_mac | okay, so apparently changing that setting in colloquy has /reduced/ the number of wrong characters spewed into the logs. we now just get a single a-circumflex instead of a stream of stuff. | 15:58.44 |
| robin_watts : I think, from the assumption that your resources changes only fix things and donât break anything, replacing sounds reasonable. | 15:59.23 |
Robin_Watts | jogux_mac: Ok. We can always back out the change and bring the old versions back if we need to. | 16:00.02 |
jogux_mac | nods. | 16:00.11 |
| mace is making progress on ATS btw, but we are behind where Iâd hoped to be :( | 16:01.13 |
ray_laptop | jogux_mac: I still see the same garbage as before | 16:01.53 |
kens | jogux_mac : Miranda still shows your 'don't' with acircumflex+Euro+Trademark :-( | 16:01.54 |
jogux_mac | :( | 16:02.10 |
jogux_mac | wonders if Paul is having this same issue | 16:02.44 |
| or if he is avoiding it by just not talking much from his mac :-) | 16:02.54 |
ray_laptop | Jogux: how are you looking at the logs ? Both chrome and firefox show the garbage | 16:03.10 |
| or avoiding contractions :-) | 16:03.27 |
jogux_mac | ray_laptop : firefox definitely shows /less/ garbage for things said since I changed that setting | 16:03.43 |
ray_laptop | Jogux: not for me | 16:03.55 |
| I guess one of could "fix" the infobot filter | 16:04.41 |
| but it's in perl | 16:04.57 |
ray_laptop | touches nose | 16:05.01 |
jogux_mac | odd. my 15:26:16 line just has a single wrong character, whereas my 15:19.45 line has a 8 or so in a row. | 16:05.15 |
| hehe | 16:05.22 |
ray_laptop | now if were in awk or python or something reasonable | 16:05.22 |
kens | For me Firefox shows worse than Miranda, I see Atilde+cent+ACircumflex+Euro+Acircumflex+Trademark | 16:06.16 |
paulgardiner_mac | Whatâs a good string for testing for the problem? | 16:07.00 |
kens | Yes that one works | 16:07.13 |
jogux_mac | the irclogs donât specify a character set; that may explain differences. | 16:08.05 |
| my browser is treating them as utf8 | 16:08.26 |
paulgardiner_mac | you didnât change resources at all, right? | 16:09.18 |
| you didn?t ? | 16:12.01 |
| you didn?t | 16:12.44 |
| Hmmm | 16:12.51 |
| You didn?t | 16:15.33 |
paulgardiner | jogux_mac: did you find a fix for Colloquy? | 16:17.31 |
jogux_mac | nope :( | 16:17.37 |
| can not find anyone else complaining about it either, which is odd | 16:17.48 |
kens | paulgardiner : that's a lot better | 16:19.12 |
| just one ? | 16:19.16 |
mace | jogux_mac: i've got 5 vm's booting | 16:20.10 |
jogux_mac | mace : woot! | 16:21.12 |
Robin_Watts | mace: 5 *useful* VMs? :) | 16:21.45 |
mace | ats master is one of them :) | 16:22.09 |
jogux_mac | mace : booting and you can log in to them? :) | 16:22.41 |
mace | so far that's 1 out of 5 | 16:23.03 |
jogux_mac | hehe | 16:23.13 |
mace | suspect some of them are static ip'd | 16:23.14 |
| looks like the win2003 one booted ok too | 16:23.46 |
jogux_mac | the windows ones at least looked to be dynamic ips | 16:23.53 |
mace | can't connect to it atm as no rdesktop | 16:23.55 |
henrys | jogux_mac: I use encoding utf-8 is that what you had before ascii? | 16:24.21 |
jogux_mac | henrys : yep | 16:24.31 |
| henrys : can you type the contraction of âdo notâ for us all then ;-) | 16:25.02 |
henrys | don't | 16:25.16 |
jogux_mac | hm. thatâs absolutely fine. | 16:25.24 |
Robin_Watts | That's a different apostrophe. | 16:25.30 |
jogux_mac | are you on mac os 10.9? | 16:25.41 |
henrys | 10.8.5 | 16:25.56 |
Robin_Watts | Jogu's apostrophe is more curly than henrys. Like â. | 16:25.58 |
henrys | Robin_Watts: UK thing? | 16:26.10 |
jogux_mac | henrys : I wonder if mac version is the thing then. Iâm on 10.9. | 16:26.26 |
Robin_Watts | not that I'm aware of. | 16:26.28 |
henrys | looks for the character viewer | 16:27.42 |
| used once every 5 years | 16:27.50 |
kens | Robin_Watts : you just did the funny character thing | 16:28.24 |
Jogu_mac2 | don't! | 16:28.35 |
| so. limechat is fine. | 16:28.41 |
kens | Sucess! | 16:28.43 |
spanners | is intrigued | 16:29.24 |
Robin_Watts | kens: Cut and Paste :) | 16:29.40 |
kens | aha :-) | 16:29.44 |
spanners | Jogu> If you right click on the input field is Substitutions > Smart Quotes (etc.) enabled? | 16:30.06 |
Jogu_mac2 | only thing enabled is 'smart copy/paste' | 16:30.27 |
kens | thinks it looks like a smart quote | 16:30.31 |
spanners | Is 'Text Replacement' on? | 16:31.13 |
jogux_mac | don't! | 16:31.15 |
kens | that worked | 16:31.20 |
jogux_mac | yay | 16:31.23 |
| so it's system preferences -> keyboard -> text -> "use smart quotes and dashes". | 16:31.36 |
kens | So what was it ? | 16:31.37 |
| ah... | 16:31.41 |
jogux_mac | which is system wide. | 16:31.50 |
spanners | ew | 16:32.41 |
jogux_mac | ⺠| 16:32.49 |
ray_laptop | jogux_mac: it didn't like your smiley | 16:36.41 |
jogux_mac | yeah, not surprised. That was just for sillyness ;) | 16:37.05 |
| robin_watts : ignoring the font stuff, your tree looks fine to me. I've not looked at the solution stuff. | 16:38.55 |
ray_laptop | Robin_Watts or mvrhel_laptop: can one of you have a look at the 4-bit fast thresholding patch on my repo | 16:40.08 |
paulgardiner_mac | don't | 16:40.13 |
Robin_Watts | jogux_mac: OK. I shall push all up to the solution then. | 16:40.40 |
kens | paulgardiner : was fine | 16:40.58 |
jogux_mac | robin_watts : tbh, I more meant "I'm uninterested in looking at the solution stuff" :-) | 16:41.11 |
paulgardiner | ah good | 16:41.13 |
jogux_mac | maybe I should create some enthusiasm for it. | 16:41.25 |
Robin_Watts | jogux_mac: Perhaps paulgardiner might have more enthusiasm for looking at the solution stuff? | 16:41.48 |
jogux_mac | hopes so | 16:41.57 |
ray_laptop | Robin_Watts, mvrhel_laptop: I know it's off topic since it isn't SOT related ;-) | 16:42.09 |
Robin_Watts | or I will accept "Whatever" as being a passed review :) | 16:42.13 |
| ray_laptop: I will look in just a mo. | 16:42.22 |
ray_laptop | Robin_Watts: OK. no rush. | 16:42.38 |
| I just want to get it in so I can send an update to potential customer | 16:43.02 |
jogux_mac | paulgardiner : can we rebase your tree on top of golden and get that in too? | 16:44.35 |
| (you might need to remove robin's font stuff) | 16:44.52 |
paulgardiner | jogux_mac: which? on master or on the smart-products branch? | 16:45.30 |
jogux_mac | paulgardiner : I think smart-products branch is good to go actually; master needs rebasing/the font stuff removed ? | 16:46.05 |
paulgardiner | Has robin_watts's stuff gone to golden then? | 16:47.19 |
jogux_mac | looks like it to me (expect the font / vs solution) | 16:48.04 |
| except, even. | 16:48.29 |
paulgardiner | Oh okay. I was planning on moving my stuff across as soon as robin_watts was done | 16:49.20 |
vecero | when using the fitz draw device, when do draws complete? is it needed to use the cookie stuff to find out, or is the pixmap mutated as calls to e.g. fz_stroke_text happen? | 16:52.54 |
jogux_mac | paulgardiner : might as well go for it now I think. (I was presuming you would take across my stuff too, but did you want me to move mine?) | 16:55.43 |
kens | Robin_Watts : tor8 see question from vecero above ^^ | 16:57.06 |
Robin_Watts | vecero: For simple files, draws complete as things like fz_stroke_text happen. | 16:57.53 |
paulgardiner | I can take both across... although not just at the moment because I'm bashing at Airprint support for the MuPDF iOS app | 16:58.09 |
Robin_Watts | But for files with transparency groups, we draw into new pixmaps for those transparency groups. | 16:58.17 |
| and when the group completes, then we blend it back. | 16:58.24 |
| so the 'main' pixmap won't be updated until we have popped all the transparency groups off the transparency stack. | 16:58.49 |
kens | didn't know that, but it makes sense | 16:59.00 |
vecero | ok | 16:59.05 |
Robin_Watts | hence you could easily have a file that showed absolutely nothing until the very last thing was drawn, then it would all appear magically. | 16:59.23 |
| The purpose of the cookie is so that 1) you can abort a render nicely from another thread. | 16:59.45 |
| 2) so you can show a progress bar. | 16:59.53 |
| 3) so you can collect details of whether any errors were ignored during the rendering process. | 17:00.09 |
vecero | is it sane to try to run that machinery "manually"? | 17:00.09 |
Robin_Watts | 4) anything else we think of later. | 17:00.18 |
| vecero: You mean to force transparency stacks to flush periodically? | 17:00.41 |
kens | thinks that owuld result in incorrect rendering | 17:00.59 |
Robin_Watts | No, that would utterly break rendering because the flushing of the transparency stack is not idempotent. | 17:01.06 |
| kens: You are right. | 17:01.20 |
vecero | hmm | 17:02.36 |
| what owns the transparency stack? | 17:03.12 |
paulgardiner | Airprint doesn't seem to find my cups-served printer. Is there some sort of config required? | 17:03.43 |
vecero | would the cleanest thing to do be create a text document impl rather like the image document impl? | 17:08.13 |
kens | vecero, perhaps if you were to explain your actual problem, rather than asking questions, it might be possible to help you | 17:08.42 |
jogux_mac | paulgardiner : I think you need a fairly recent cups perhaps | 17:09.01 |
| paulgardiner : when I upgraded to 1.5.3 it automatically started appearing in airprint | 17:09.27 |
paulgardiner | I update my server every time a new version of opensuse comes out | 17:09.40 |
jogux_mac | ah, I'm on debian, that may also matter. | 17:09.49 |
paulgardiner | I'm on 1.5.4 :-) | 17:10.05 |
jogux_mac | I didn't even know suse was still around :-) | 17:10.06 |
vecero | kens: I'm trying to draw a line of text to a pixmap, in order to show in the status bar in the Wayland viewer | 17:10.32 |
Robin_Watts | vecero: Sorry, got distracted. | 17:10.47 |
kens | THat's greek to me, but maybe Robin_Watts will understand | 17:10.56 |
paulgardiner | I might give up for now and look at rebasing those commits | 17:11.28 |
Robin_Watts | vecero: If I'm understanding you correctly you're looking to use MuPDF to render a line of text into a pixmap? | 17:11.51 |
jogux_mac | paulgardiner : you may need avahi-daemon installed too | 17:12.30 |
Robin_Watts | The transparency stack is owned by the draw device. | 17:12.32 |
vecero | Robin_Watts: yes | 17:12.53 |
Robin_Watts | Or more generally, every device needs to implement as much or as little of a transparency stack as is required. | 17:13.12 |
| the text extraction device doesn't need a transparency stack, for example. | 17:13.33 |
paulgardiner | jogux_mac: yeah that doesn't seem so easy to find under suse | 17:13.50 |
Robin_Watts | So, certainly there is nothing stopping you opening a draw device on a pixmap, then calling the render text entrypoint in the device to get the text drawn. | 17:14.34 |
| Then close the device. | 17:14.44 |
| In that case, there is no transparency stack in play. | 17:14.52 |
| so the text will come out instantly. | 17:15.00 |
vecero | so in order to draw the status text (e.g. the "search" prompt), I need to draw just a simple line of text, and since fitz is already there and capable I'm using it to render a single line of text :P | 17:15.15 |
Robin_Watts | and the cookie etc is all irrelavent. | 17:15.16 |
vecero | okay | 17:15.17 |
| thanks for the help, I'll write some more code and see how things turn out | 17:15.39 |
Robin_Watts | vecero: The problem with that, is that once you've drawn the 'search' prompt into the pixmap, you can't easily erase it. | 17:15.51 |
| Well, you could draw a white rectangle or something, and then put the text into that. | 17:16.10 |
vecero | I'll probably do that | 17:20.43 |
paulgardiner | robin_watts have you finished pushing to sot/golden for now? | 17:24.11 |
| jogux_mac: actually, it looks like all my commits on master are subsumed into robin_watts's now. There are only yours to push. Is that right? | 17:26.27 |
| rebase and push mean | 17:26.50 |
jogux_mac | really? I thought you'd done more than that :-) | 17:27.05 |
| you definitely have some to go on the release branch right? | 17:27.41 |
paulgardiner | Oh yes. Quite a few. | 17:28.02 |
| I can just push that, right? We've reviewed each others already | 17:28.21 |
jogux_mac | I agree about master | 17:28.32 |
| and yes, just push release branch | 17:28.40 |
| you put the tags back on the release branch in the end right? | 17:28.52 |
paulgardiner | Oh I do have some private subbranches of master that have stuff we might want | 17:30.05 |
| tags: no, but I was just about to | 17:30.30 |
Robin_Watts | paulgardiner: For now, yes, all done. I'll have fonts ready to go soon, I hope, but that can wait til after your lot. | 17:31.20 |
paulgardiner | robin_watts here it looks like golden/master is 2 back from the VS-solution commit | 17:32.58 |
Robin_Watts | paulgardiner: Damn. Pushed the wrong SHA. 2 mins. | 17:33.29 |
| pushed. | 17:34.15 |
paulgardiner | ta | 17:34.30 |
kens | Goodnight all | 17:49.06 |
Robin_Watts | Jogux, paulgardiner: OK, 4 commits on robin/master that should finally get me up to date. | 17:50.35 |
paulgardiner | I've just got some rebased | 17:51.03 |
Robin_Watts | paulgardiner: Let me look at yours now. | 17:51.15 |
jogux_mac | Robin_Watts : putting icky font stuff up for review at 10 to 6 on a Friday seems.... optimistic :-) | 17:51.20 |
| or very careful timing. | 17:51.36 |
Robin_Watts | jogux_mac: Ask your self a question. Do you feel Bazely, punk? | 17:52.16 |
jogux_mac | hehe | 17:52.26 |
paulgardiner | pushed to my repo | 17:53.34 |
| jogux_mac: you may want to look too in case I've broken one of yours | 17:53.56 |
| robin_watts I'm not sure how the first of pete's interacts with your "manifest" commit | 17:54.29 |
Robin_Watts | paulgardiner: Your 3 commits look fine to me (on paul/master). Looking at petes now. | 18:00.00 |
| oh, they are josephs :) | 18:00.12 |
| So pete is sidestepping the link-with-manifest except in the 2005 case. I think that's wrong. | 18:01.21 |
paulgardiner | You could rebase your stuff just before those and then we can worry about that tomorrow | 18:02.20 |
Robin_Watts | otherwise it looks fine to me. | 18:03.44 |
| In petes last one, shouldn't the Url_destroy(); be 2 lines higher to prevent a leak in an error case ? | 18:05.09 |
mvrhel_laptop | bbiaw | 18:49.26 |
Robin_Watts | marcosw_: You here? | 21:00.38 |
marcosw_ | Robin_Watts: wasn't at 2:00 but am now. | 22:15.31 |
Robin_Watts | marcosw_: Still here? | 23:02.46 |
marcosw_ | I am. | 23:02.53 |
Robin_Watts | marcosw_: If you visit twiki.ghostscript.com and register, I can enable you on the GhostDocs web. | 23:03.25 |
| And that'll be everyone I think. | 23:03.34 |
marcosw_ | stalking me on irc was easier than sending an email? :-) | 23:05.20 |
| done | 23:05.34 |
Robin_Watts | marcosw: yup. | 23:06.17 |
| ah. I've been asking people to use their full names. | 23:06.41 |
marcosw_ | Sorry, didn't realize. Can I edit this or do I have to re-register? | 23:07.58 |
Robin_Watts | Reregister and I'll delete the old you. | 23:08.10 |
marcosw_ | done. | 23:10.18 |
Robin_Watts | You should be able to see the GhostDocs web. | 23:12.52 |
marcosw_ | still saying access denied. I'll try login out and back in. | 23:17.25 |
| that worked. | 23:18.11 |
Robin_Watts | fab. | 23:18.21 |
| At the moment, all the webs are public except for GhostDocs. | 23:18.44 |
| So feel free to put anything on there that you want. | 23:19.00 |
| I can even make other webs if you want (if we want for one for cluster or something) | 23:19.15 |
marcosw_ | okay, thanks. | 23:20.10 |
| Forward 1 day (to 2014/03/29)>>> | |